The biological, medical and social consequences of HIV infection of the central nervous system (CNS) include neurodegeneration, disability from neurocognitive disorders, increased mortality, and decreased quality of life. The CNS may serve as reservoir and sanctuary for HIV because the blood-brain barrier shields infection from certain immunologic and pharmacologic defenses. Recently, quantitative assays for HIV RNA and chemokines have been developed that promise to enhance our understanding of HIV neuropathogenesis and to facilitate the diagnosis and treatment of HIV CNS infection. This proposal outlines an integrated series of clinical and laboratory investigations to assess the role of the CNS in the development of antiretroviral (ARV) resistance and to evaluate strategies for therapeutic intervention in humans through the study of cerebrospinal fluid (CSF). Three clinical studies form the core of this 5-year proposal. Study 1A is an open-label trial of individualized combination ARV therapy (ART) for HIV neurocognitive disorder. The principal objective is to correlate improvements in neurocognitive function with CSF virologic (HIV RNA) suppression and CSF chemokine changes during ART. Eighty participants with neurocognitive disorders will be enrolled for a total of 155 treatment-observation periods of 12-weeks each. Study 1B is a controlled comparison with virologic responses in CSF and plasma with ART in individuals without neurocognitive disorders. Its main goal is to compare rates of change in CSF and plasma HIV RNA levels (HIV dynamics) and chemokines during ART, and to evaluate determinants of virologic efficacy and failure in the CSF. Fifty neurocognitively normal participants will be enrolled for 80 treatment-observation periods of 12- weeks each. Studies 1A and 1B have a parallel design in which participants are randomized to one of two treatment strategies (plasma versus plasma+CSF) based on the use of phenotypic ARV resistance testing. Study 2 compares virologic and immunologic (chemokine) responses in CSF and plasma during the withdrawal of antiretroviral treatment ("structured treatment interruption") in 2o subjects. Its goal is to determine the nature and extent of compartmentalization of CSF responses. Intensive laboratory investigations of HIV genetics , chemokines and chemokine receptors genotypes will be applied to the fluid samples derived from these studies. Performance of the work proposed here will benefit from this grant's relationship to the HIV Neurobehavioral Research Center and to the infrastructure that the HNRC provides in neurocognitive assessment, participant recruitment, data management, statistics and tissue archiving.